Possible bug in QUADPACK routine

[Paul Barclay <pbarclay at caltech dot edu>]

Hi,

I installed GSL using the provided RPMS of my PIII laptop running Redhat
7.0.

I've just begun using GSL, and so far I've only tried out some of the
quadrature routines (qgsl_integration_qags and gsl_integration_qagiu).
Numerically they seem fine, but they have a nasty habit of causing a
segmentation fault + core dump rather then return an error code whenever
they are not able to do the integral.

This small problem asside it's a great library and I look forward to
watching it evolve.  Hopefully I will never have to worry about converting
fortran to C again!  One addition which would be very valuable I think are
Bessel functions capable of taking complex arguments.  Currently the
fortran AMOS implementation available on Netlib is the only free code I've
been able to find to do this.

Thanks,

Paul Barclay